PsyAsia Live at SIOP: PsyAsia’s consultants are currently attending the SIOP Conference in Chicago and posting brief summaries of the most interesting talks they attend. Here is one of them!
The start to this presentation was interesting as the presenter pointed out how online psychometric testing was like a train that had left the station and we as psychologists were playing catch up. That’s very true, especially in the case of unproctored (unsupervised) testing. There was a push from industry to provide unsupervised tests. Some test publishers pushed ahead and only considered the ramifications later, others held back and only allowed their tests to go online and unproctored once they had seen what happened to others. In addition, there have of course been new publishers since the dawn of online testing and these guys have had the benefit of being able to develop their tests with internet assessment in mind.
The presentation focussed more on ability tests and the issue of cheating. A psychologist from Cox pointed out that their policy is to trust applicants even at the early stages given that they have a culture of trust within the organization. So, the question is, are they right in being so trusting or can candidates cheat tests in some way. Another panellist pointed out that if cheating was occurring there should theoretically be different score distributions for the cheaters versus the non-cheaters and in fact in his research he was not able to find any significantly different score distributions. From the evidence presented it did appear that as a whole, candidates do take the assessment processes seriously and do provide accurate data.
However, one does need to keep in mind that the results were based on research with a number of groups of individuals and that in critical roles, if just one person cheats and the decision-maker is unaware of this, there will be negative implications. Research is showing us that cheating is not a worrisome issue with unproctored tests and that is good news. Yet what about that one person, once in a while, who asks somebody else to complete his online aptitude test on his behalf? If this person makes it through the remainder of the recruitment process, one can expect problems later. This of course is costly for the organization.
The best approach then is perhaps to take comfort in the fact that most candidates appear to be honest in the unproctored psychometric testing process, whilst keeping in mind that the odd person may make it through whilst being dishonest and undetected. As per the guidelines of the International Test Commission, it is therefore still a useful policy to re-test any candidate with a supervised, parallel form of the same test at the final stages of the recruitment/selection process.